Improving fair value accounting
JOSÉ VIÑALS
Deputy Governor
Banco de España
The turmoil on international financial markets is proving complex to a degree that few could have
anticipated when it initially emerged in the summer of 2007. There is still much uncertainty over the duration
and potential impact of this turmoil on the real economy.
This episode has revealed a series of flaws in various areas of the international financial system.
One issue on which regulators, supervisors and other interested parties are focusing is the application
of fair value. In many cases discussions turn on quantitative and qualitative matters geared to improving
valuation methods and their implementation, especially when applied to complex financial instruments.
There is also in-depth reflection about what information institutions should provide investors regarding
the application of fair value, so that investors may take well-grounded decisions.
Another area of the debate on fair value considers to what extent its application affects management
and investment decisions, and particularly how it may exacerbate procyclical behaviour by financial markets.
To examine the relationship between valuation and procyclicality and to identify some solutions to the
perverse interaction of the two, the article discusses the advantages of fair value and its limitations, stressing
in particular some of the most relevant ones which have emerged during the current financial turmoil.
In addition, it puts forward some ideas that might contribute to improving fair value: the use of reserve
valuations and of dynamic provisions. It is argued that they can not only improve fair value accounting
but also lessen financial procyclicality.
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
121
ARTICLES
José Viñals: “Improving fair value accounting”
T
he turmoil on international financial markets
is proving complex to a degree that few could
have anticipated when it initially emerged in
the summer of 2007. There is still much uncertainty
over the duration and potential impact of this turmoil
on the real economy.
This episode has revealed a series of flaws in
various areas of the international financial
system. One topic on which regulators, supervisors
and other interested parties are focusing is the
application of fair value. In many cases discussions
turn on quantitative and qualitative (governance)
matters geared to improving valuation methods
and their implementation, especially when
applied to complex financial instruments. There
is also in-depth reflection about what information
institutions should provide investors regarding the
application of fair value, so that investors may take
well-grounded decisions.
Another area of the debate on fair value considers
to what extent its application affects management
and investment decisions, and particularly how it
may exacerbate procyclical behaviour by financial
markets. The procyclical behaviour of financial
agents means that in good times they tend to increase
their risk-taking, and financial vulnerabilities
thus build up. These vulnerabilities become
manifest when the economic cycle turns adverse,
prompting changes in market participants’ strategies
and amplifying the cycle.
Procyclicality is certainly an intrinsic characteristic
of financial markets. But beyond a certain point it
arguably generates highly adverse effects on long-term
growth prospects (“excessive procyclicality”). This is
the case especially when it leads to or exacerbates
myopic behaviour and excessive short-termist
behaviour on the part of financial institutions.
The adverse consequences of procyclicality tend
to be more intense at times, as at present, when
the interplay between asset valuation and leverage
has become more important. If valuation methods
tend to introduce incentives to increase leverage
and mispricing risks, the adjustment process
when economic conditions change will be more
pronounced, amplifying its adverse impact on the
economy. As it is known, leverage and mismatches
tend to grow slowly, but downward adjustments
occur rapidly.
122
It would therefore be desirable to introduce incentives
to mitigate this procyclical behaviour, and in any event
to avoid regulations that might encourage it. Indeed,
during the conception of Basel II, procyclicality was
a controversial and widely debated topic, and the
Basel Committee itself included various mitigating
mechanisms in the Accord. On the contrary,
and surprisingly, there has been very little debate
until recently concerning the new accounting rules
financial institutions have to apply.
The use of fair value to approximate the value of
financial instruments is nothing new. Its use has
gradually increased over time and has gained in
significance as specific products, such as derivatives,
have become widespread. For these and other
financial instruments, valuation at cost has been
inadequate in providing a true and fair view
of companies. Historical cost is less informative
from a time perspective, and is also insensitive
to the signals market prices emit, which hampers
agents’ decision-making and market discipline.
These limitations of cost-based accounting support
the use of fair value as the valuation method on which
accounting rules should be based. Fair value offers
evident advantages in that it provides for a better
approximation to the economic reality prevailing at
each point in time.
That said, and as will be argued in this article, fair
value would need to be perfected in several areas.
In particular, it should allow a better evaluation
of risks and profits over the business cycle and
it should be compatible with promoting financial
stability. The current financial turmoil dating back
to the summer of 2007 exemplifies the problems
of the present conception of fair value, especially
when there are no deep and liquid markets (or these
are disappearing), and the products being valued
are complex.
When markets disappear, and it is no longer possible
to use a market price to value financial instruments,
it becomes necessary to resort to internal valuation
models based on inputs not directly observable in
the market and which, in turn, are subject to high
procyclicality. In this way, a negative dynamic
arises and intensifies the problems as a result of the
reaction of investors and bank managers, who act to
limit losses.
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
ARTICLES
José Viñals: “Improving fair value accounting”
To examine the relationship between valuation and
procyclicality and to identify some solutions to the
perverse interaction of the two, the rest of this article
is structured as follows. The first section discusses
the advantages of fair value and its limitations,
stressing in particular some of the most relevant
ones which have emerged during the current
financial turmoil. The second section analyses
factors that might contribute to improving fair value
since, as already mentioned, a return to historical
cost seems neither viable, possible nor desirable.
The final section draws the main conclusions
and policy implications.
1| ADVANTAGES AND LIMITATIONS
OF FAIR VALUE
WHAT HAS THE FINANCIAL
TURMOIL TAUGHT US?
Like any other valuation method, fair value
cannot be considered to be a perfect description
of reality. Accordingly, to assess whether fair value
is appropriate, there are two pertinent questions.
First, does it offer suitable (reliable, comparable
and relevant) information enabling economic
agents to take well-grounded investment decisions?
And second, what weaknesses does it entail?
1|1 The advantages of fair value
Starting with the advantages, it should be noted that
financial innovation in recent decades has developed
a series of products for which valuation at cost has no
use. A clear example of this are financial derivatives,
where fair value has proven to be the only method
capable of offering a transparent, relevant and
reliable valuation. However, the growing complexity
of some of these products, and certain limitations in
valuation models, have proven problematic recently.
I shall return to this point later.
More broadly, fair value provides additional
advantages in that it offers a closer view of the
actual situation of financial markets. In other
words, and unlike cost-based methods, it enables
the information contained in market prices at each
point in time to be included, which is useful for
those who have to commit funds, or have funds
committed, to a financial institution. Likewise,
this information is closer to that which institutions
themselves use for management purposes,
which contributes to introducing appropriate
incentives between managers and investors.
Consequently, fair value is associated with
greater market discipline, since the action of
market players will have a more direct bearing on
institutions’ decisions. What is more, insofar as
investors consider the information in financial
statements to be useful and relevant, confidence will
be reinforced. Both these matters, greater market
discipline and strengthened confidence, entail
improvements in terms of efficiency.
Fair value, as indicated, is a market estimation of
financial instruments, which is what enables it to
include ahead of time all the information available at
a given moment. It will thus contribute to detecting
potential solvency problems that may be liable
to affect institutions, as it will rapidly reflect any
deterioration in the quality of their balance sheets.
In sum, the advantages of fair value, and moving
beyond the presence of a series of increasingly
relevant financial instruments for which valuation
at cost is not viable, have to do with improvements
in the allocation of resources that internalise the
information present in financial markets.
1|2 The disadvantages of fair value
Nonetheless, and as stated, no valuation method is
free from limitations, as methods are no more than
conventions on how to measure value.
The traditionally disputed limitations of fair value
can be grouped around three ideas: the subjectivity
of valuation methods; the greater volatility induced;
and, in connection with the latter, the excessive
emphasis on the short term.
When financial markets are active, the prices traded on
them conceivably reflect consensus between buyers
and sellers about the future cash flows of financial
instruments, and on the degree of uncertainty
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
123
ARTICLES
José Viñals: “Improving fair value accounting”
surrounding them. Under these circumstances, fair
value coincides with market price.
short-term perspective motivated by the changes in
accounting value reflected in financial information.
However, there are not always active markets,
and prices are therefore not always available
for use in applying fair value. In these
circumstances, estimating fair value involves the
use of valuation methods that allow estimation of
what the market price of the financial instrument
would be. Irrespective of the model or technique
used to develop valuation models, there is some
subjectivity in their design. Expressed otherwise,
valuation will be affected by the judgement of
those who have to develop a model, since they have
to take decisions on the theoretical grounds they
apply and on the assumptions and simplifications
they consider necessary.
The greater sensitivity of fair value to financial
market circumstances may affect the behaviour
of bank managers, encouraging undesirable
behaviour in terms of appropriate risk management.
Thus, for example, if a revaluation of assets was
reflected in the institution’s profits, that might
lead to an increase in the dividend pay-out to
shareholders, restricting institutions’ capacity
to soften intertemporal shocks.
Associated with this subjectivity is so-called model
risk, i.e. the probability that errors will be made in
the valuation of a specific instrument owing to the
use of inappropriate techniques or to having made
assumptions that prove unsatisfactory.
Furthermore, the use of valuation models may
give rise to new information asymmetries, creating
moral hazard problems. When there are incentives
to manipulate the information reported to the
markets, the use of valuation models opens up the
possibility –which is more evident than when using
historical cost– of institutions cherry-picking specific
parameters or assumptions, thereby accentuating
information asymmetry problems.
Secondly, it has been argued that fair value increases
volatility on bank balance sheets, on profit and
loss accounts, and consequently on the levels
of regulatory capital that banks must hold.
As described below, this would intensify the
procyclicality of financial markets.
Insofar as fair value takes into consideration market
conditions at a specific time, the profit and loss
account would be excessively influenced by these
potentially very temporary market conditions.
This argument would be weightier if the volatility
observed on markets were not in response to
fundamentals but to spurious reasons. What
is more, this volatility might be exacerbated
by investors’ decisions if they were to act from a
124
If results were affected by greater volatility,
this would be transmitted to regulatory capital,
contributing to procyclical behaviour by bank
managers. The contractionary phase of the cycle
may thus be accompanied by an across-the-board
fall in valuations, which would feed through to
results and to institutions’ capacity to generate
reserves. This effect would compound the greater
requirements attributable to credit risk that are
typical in these contractionary phases, insofar as
defaults are negatively correlated to GDP. And
this at a time in the cycle when institutions would
face greater difficulties increasing their own funds
through fresh capital. Accordingly, with a view to
meeting minimum regulatory capital requirements,
bank managers might react by reducing the credit
they grant to the real economy, which would entail
a highly adverse impact.
Thirdly, and as a result of the greater volatility
induced in banks’ profits and loss accounts, the use
of fair value might create perverse incentives in
banks’ management decisions, placing excessive
emphasis on the short term. These perverse incentives
may be specifically perceived in investment
decisions, e.g. avoiding those sectors in which the
volatility of credit ratings is greater, but they will also
affect managers when their compensation is linked
to accounting results. Managers might therefore
benefit by setting greater store by those decisions
that result in profits in the short term, without taking
sufficiently into account other long-term strategic
questions, including those relating to risks taken.
Further, in the case of relatively illiquid instruments,
managers’ short-termist decisions might ultimately
affect (or alter) the prices which fair value is actually
intended to approximate.
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
ARTICLES
José Viñals: “Improving fair value accounting”
1|3 The limitations exposed
by the financial turmoil
The current turmoil in the international financial
system has highlighted some of these limitations
of fair value.1 In fact, this is the first time that
many of the products developed recently, in
particular the most complex ones, have been tested.
This is the case, too, for the models which, under
the fair value approach, have been used to value
such products. It is therefore worth analysing in
some detail the implications that may derive from
these circumstances.
Since last summer, liquidity on many markets
has dried up considerably. Entities that valued
specific products using pre-turmoil market prices
have been obliged to resort more to mark-to-model
valuations, largely based on inputs not directly
observable in the markets (default probabilities,
correlations between the defaults of a portfolio, etc.).
In this transition process, certain fair value
implementation weaknesses have come to light.
For one thing, adaptation difficulties have proven
more acute for those institutions which, prior to
the emergence of the turmoil, resorted to a limited
number of information sources in their valuation
process to determine the prices at which to value
financial instruments, and in particular the most
sophisticated and complex instruments. For example,
some entities drew excessively on primary market
prices as an approximation to the fair value of the
products to be valued. Hence, when these markets
virtually disappear for specific business segments,
valuation problems emerge with greater intensity.
Moreover, difficulties have been greater for
those entities which, before the turmoil, had
not developed appropriate contingency plans or
which had in fact earmarked insufficient resources
to the development of fully fledged valuation
models (model development, stress testing, laying
of contingency plans, etc.).
The lack of liquidity, along with greater valuation
problems, became manifest for a set of complex
products that were rapidly developed in recent years
under the originate-to-distribute banking model.
1
In particular, these products combine and repackage,
via asset securitisation, new financial instruments
that are re-sold to investors. The successive
combinations of securitisation tranches substantially
alter the distribution of losses on these products,
so that the level of risk, which increases with each
successive securitisation, is much more difficult to
calculate. Beyond the complexities inherent in the
valuation of this type of highly sophisticated product,
the turmoil has highlighted certain shortcomings
which should be taken into consideration.
These models were designed under benign
economic conditions, and without due
consideration being given to how they would
operate if conditions turned adverse. In this respect,
the valuation models used have not included all the
relevant risk factors and, in particular, they have
suffered from the absence of three of these risks:
model risk, liquidity risk and counterparty risk.
For example, many models have not taken
sufficiently into account that the underlyings of a
lot of these complex products were US subprime
mortgage loans, which were therefore sensitive to
changes in interest rates, to house prices and to
borrowers’ incentives. The correlations between
defaults in the subprime loan portfolio behind
asset-backed
bonds
were
considerably
underestimated.
Ultimately, one element that has characterised
the current turmoil, in particular at its onset, has
been the lack of investor confidence. It was not
clear, first, who was bearing the risks nominally
considered to have been transferred, and second,
what exposures were actually committed. This lack
of transparency has also been evident in relation
to the information provided to markets about the
models applied to estimate the fair value of the
different products.
In order to properly understand financial
information, investors need to be able to judge how
financial instruments have been valued. For this
purpose, they need to know, first, the committed
exposures, and second, how they are being valued:
which techniques are applied, which inputs are
used, what assumptions have been made and what
is the degree of sensitivity of the valuation to the
emergence of different scenarios.
See, for example, CEBS (June 2008), FSF (April 2008), BCBS (June 2008) and the Banco de España FSR (04/2008).
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
125
ARTICLES
José Viñals: “Improving fair value accounting”
In sum, the financial turmoil has highlighted
significant weaknesses regarding the implementation
of fair value in at least three areas. First, from a
quantitative standpoint, it has exposed shortcomings
in the design of valuation models, which have not
properly captured the characteristics of the most
complex products. Second, from a more qualitative
perspective, it has highlighted governance problems,
in that systems have not been appropriately designed
to verify and test the valuations made. And third,
the information reported to the market does not
appear to have been sufficient to allow users of such
information to understand it.
These implementation-related limitations of fair
value have affected agents’ behaviour, offering
arguments in favour of the general limitations
indicated earlier. In particular, fair value has been
seen to induce greater procyclicality in certain
circumstances.
Indeed, investors in the most senior securitisation
tranches were not prepared to manage much
higher levels of risk than those initially considered.
These types of instruments, generally included in
the trading portfolio, rapidly began to lose value,
meaning that many investors took the decision to
sell once the value fell below specific thresholds
in the fair value of the instrument. This process
exerted downward pressure on valuations, feeding
back into further declines. The final effect has been
a very significant impact on the profit and loss
accounts of numerous institutions, many of which
have had to shore up their capital position under
difficult market conditions.
2| IS IT POSSIBLE
TO IMPROVE FAIR VALUE?
As argued so far, a return to valuation at cost
appears to be neither feasible nor desirable. Yet the
application of fair value under very adverse financial
market conditions has highlighted significant
limitations which bear negatively on financial
stability. Improving its functioning would appear
to be necessary.
This section is intended to offer some thoughts that
may contribute to the debate on how to design an
126
accounting framework capable of combining two
requirements of great importance for the financial
system: to offer relevant, reliable and comparable
information so that investors may make their
investment decisions appropriately; and, at the same
time, to contribute to financial stability, or at least
to limit the incentives which, from the regulatory
angle, may contribute to impairing stability.
In other words, improving fair value will involve
seeking valuation mechanisms that give a truer and
fairer view of the risks and profits institutions take
during the cycle. There are two potentially useful
approaches here.
2|1 Valuation reserves
Firstly, institutions could set aside valuation reserves
for those more complex structured products that
are mark-to-model. These reserves would entail the
recognition, in accounting terms, of the uncertainty
associated with the calculation of fair value under
specific circumstances.
As earlier stated, many of the problems of applying
fair value concern the very complexity of the products
to be valued, and the speed with which market
conditions may deteriorate at a given moment.
Consequently, it can be extremely difficult to make
estimates of certain inputs that are necessary in
the valuation models, and which moreover are not
(or may cease to be) observable in the markets.
Faced with these difficulties, it is worth performing
different stress tests envisaging different possible
scenarios. That provides a measure of the degree of
uncertainty surrounding the valuation of a specific
instrument at a given time. Ideally, institutions
should therefore reflect these measurements of
uncertainty in their financial statements: objective
and transparent valuation reserves might play a useful
role. For example, among other inputs valuation
models require estimates of probabilities and of
loss given defaults. Estimating these is, especially
for the more sophisticated products, complex and
may be subject to a high degree of uncertainty.
Accordingly, banks should have different estimates
of the fair value of the instrument depending on
the distinct values these inputs may take given
different assumptions and scenarios. Institutions
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
ARTICLES
José Viñals: “Improving fair value accounting”
should reflect in their accounts the estimation of the
fair value of the instrument, along with a valuation
reserve that reflects the uncertainties surrounding
this estimation.
As earlier stated, the valuation reserves should be
objective and transparent. And two further qualifying
conditions should be added here. First, they should
be symmetrical. That is to say, these reserves have
to function both in the good times of the cycle and
when economic conditions turn adverse. In other
words, it is not a question of generating buffers to
face difficulties, but of improving the valuation of
complex financial instruments by explicitly and
transparently incorporating the uncertainty that
surrounds the valuation.
Secondly, and running counter to a view widely held
at present, fair value implementation problems do
not emerge ex novo when economic and financial
conditions turn adverse; they already exist in good
times. It is precisely in these periods when agents act
with excessive optimism, valuing risks inappropriately.
Can it be disputed at present that the losses being
posted do not stem from the excesses committed in
the prior years of strong economic growth? This is why
the valuation reserves should act at all times so as to
offer a more appropriate valuation of the instruments
for which valuation models are needed.
All in all, the inclusion of valuation reserves gives a
truer measurement of the value of the instrument
than that which would be obtained from a direct
application of a point estimate of its fair value,
since it reflects explicitly in the financial statements
a measurement of the degree of uncertainty
that institutions manage when they are valuing
instruments. Accordingly, investors may take their
decisions on a firmer basis.
2|2 Dynamic provisions
The second mechanism that might help improve fair
value is dynamic provisions. These are, in fact, a good
measurement of the fair value of a loan portfolio.
2
3
Conceptually, provisions should be understood as
adjustments to the book value of the loans on deposit
institutions’ balance sheets. These adjustments take
into consideration the impact that the credit risk
borne by the institutions entails for these loans.
Debate turns on whether this correction in value
should envisage exclusively the specific losses that
may be identified for specific loans as of the date of
the financial statements, or whether consideration
should also be given to the losses that the institution,
aware that it has incurred them, cannot specifically
identify at the level of each individual loan.
In other words, is it reasonable that investors should
be surprised in the future by what is already known is
going to happen, or is it better that they should have
this information when they are about to take their
decisions? What better contributes to the true and
fair view of banks reporting financial information?
In the field of credit risk, saying that “it is known what
is going to happen” is no more than verifying a fact
highlighted both by the supervisory experience and
by the economic literature. Drawing on supervisory
experience, several episodes of financial instability
have shown that, following periods of strong credit
expansion, the risks that have built up materialise
in the form of defaults. It seems difficult to counter
the argument that the losses affecting a good number
of international banks today respond to the excesses
committed by their managers during the long
expansionary phase preceding the financial turmoil.
The economic literature, from an empirical viewpoint,
has also found firm evidence of a positive relationship
between processes of rapid credit growth and losses
attributable to credit risk,2 which emerge with a certain
lag. Theoretically, various arguments have been used
relating to information asymmetry and to problems
of incentives, which justify this relationship between
credit growth and default.3 In other words, at good times
in the business cycle risks tend to be underestimated,
and build up in bank balance sheets.
Therefore, there are objective, well-founded
reasons both in the supervisory experience and in
See Jiménez and Saurina (2006).
During economic booms bank managers tend to underweight the possibility of bad borrowers being financed (the opposite happens during recessions). The literature
offers several explanations to rationalise fluctuations in credit policies, such as herd behaviour (See Rajan, 1994), agency problems (See Williamson, 1963) and the
so-called institutional memory hypothesis (See Berger and Udell, 2004). Borio (2007) summarises the main factors that explain why during good times it is possible
to observe an overextension in risk-taking.
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
127
ARTICLES
José Viñals: “Improving fair value accounting”
economic theory to justify the fact that a correct
valuation of loans in the credit portfolio should
also include the factors of risk which, though they
have not been specified at the level of individual
loans, can in fact be quantified through statistical
procedures. The absence of value corrections
for these reasons distorts the true and fair view
that financial information should provide,
hampering decision-making by investors. Dynamic
provisions, as an approximation to the fair value
of banks’ loan portfolios, mitigate the fact that, at
favourable times in the cycle, risks are assumed
and build up but are only disclosed with a delay
in financial institutions’ profit and loss accounts.
With regard to improving fair value through the
two methods proposed (valuation reserves and
dynamic provisions), it should be borne in mind that
two instruments are involved that would be totally
objective and transparent for investors, so that the
latter may have the fullest information possible.
There are two further arguments in favour of valuation
reserves and of dynamic provisions. The first takes
into account the fact that the information reported
in the financial statements should ideally converge
with that which institutions use for management
decision-making. Conceivably, managers will
have to consider what uncertainty surrounds their
valuations for the different financial instruments; it
may likewise be assumed that when they analyse
the credit quality of their balance sheet, they will
bear in mind the best estimate available of credit
risk losses. If these two assumptions are correct,
improving fair value in this direction will contribute
to bringing the information analysed by managers
and investors into line. Should managers fail to take
into account risks and profits over the cycle in their
decision-making, the introduction of the proposed
changes to fair value will help them do so. And that
is good for risk management while, at the same time,
investors are offered better information.
The second of these arguments in favour of
valuation reserves and of dynamic provisions is
that mechanisms are involved that not only offer
a truer and fairer view of the bank, but which also
contribute to mitigating the current procyclicality
of accounting rules, and in general the procyclical
128
behaviour proper to financial market participants.
That is to say, both measures favour a longer-term
view, and in particular one “through the cycle”, both
for managers and for those who need to evaluate the
risk profile of banks.
2|3 General principles for the review
of fair value
Beyond the specific instruments that may be
considered, the significance of which has been
outlined, there is a series of general guiding
principles that are relevant when reviewing fair
value. These are listed below.
SEARCH FOR BETTER VALUATIONS
This principle should be binding both in the more
quantitative aspects referring to valuation models
(consideration of all the necessary factors of risk; use
of tried and tested methodologies, etc.) and as regards
providing a more realistic view of the risks and
profits associated with business activity throughout
the cycle (e.g. through dynamic provisions and
valuation reserves).
MINIMISE THE PROCYCLICAL IMPACT INDUCED
BY FAIR VALUE
It should be acknowledged that, during the favourable
phases in the cycle, financial market participants
display a natural behaviour that leads them to
build up a series of risks which, when the situation
turns adverse, tend to materialise. Accounting rules
should not encourage this type of behaviour. From
the standpoint of the application of fair value, the
use of dynamic provisions and of valuation reserves
may contribute to mitigating such risks.
GOVERNANCE
Both quantitative and qualitative aspects are relevant
in the implementation of valuation models. Sufficient
resources should be set aside, and models should be
revised and tested by independent units and subjected
to different stress tests, among other considerations.
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
ARTICLES
José Viñals: “Improving fair value accounting”
TRANSPARENCY
So that investors may take well-founded investment
decisions, they need to be able to receive the
necessary information. It will not only be important
to know the volume of exposures valued by means of
fair value, but also matters relating to the valuation
method and inputs used, to the assumptions made and
to those aspects that may contribute to users’ better
understanding of the financial information.
This article proposes a debate on the suitability of fair value as a valuation method, in the light
of the dysfunctions that the current financial turmoil has highlighted.
The main argument holds that accounting rules relating to asset valuation involve the application
of fair value as the best standard for providing final users of financial statements with relevant, reliable and
comparable information. However, it is proposed that fair value needs to be improved in order to strengthen
the relevance, reliability and comparability of information. This can be done through two specific mechanisms
(dynamic provisions and valuation reserves) within the framework of some more general guiding principles
for this review process.
Regarding the two specific mechanisms discussed, both share important characteristics that are relevant
in terms of their effectiveness and usefulness for market participants: objectiveness, transparency
and symmetry throughout the business cycle. As to the general principles, these are: the search
for better valuations; minimising the procyclical impact induced by fair value; strengthening governance;
and promoting transparency.
Importantly, at the level of financial institutions, the proposed change in fair value should enable the
quality of the information understood in the aforementioned terms to be compatible with the application
of best practices in risk management, in order to restrict the generation of inappropriate incentives
that may impair the stability of the financial system. While fair value is a step forward when compared,
for instance, with cost valuation, the current turmoil very clearly shows that it is by no means the end
of the road. Let us continue along it.
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008
129
ARTICLES
José Viñals: “Improving fair value accounting”
BIBLIOGRAPHY
Banco de España (2008)
Financial Stability Report, April
Basel Committee on Banking Supervision (2008)
“Fair value measurement and modelling: an
assessment of challenges and lessons learned from
the markets stress”, June
Berger (A.) and Udell (G.) (2004)
“The institutional memory hypothesis and the
procyclicality of banks’ lending behaviour”, Journal
of Financial Intermediation, No. 13 (4), 458-95
Borio (C.) (2007)
“Change and constancy in the financial system:
implications for financial distress and policy”,
BIS Working Paper, No. 237
Financial Stability Forum (2008)
“Report of the financial stability forum on enhancing
market and institutional resilience”, April
Jiménez (G.) and Saurina (J.) (2006)
“Credit cycles, credit risk and prudential regulation”,
International Journal of Central Banking, vol.2, No. 2,
65-98
Rajan (R.) (1994)
“Why bank credit policies fluctuate: a theory and
some evidence”, Quarterly Journal of Economics,
No. 109 (2), 399-441
Williamson (O.) (1963)
“Managerial discretion and business behaviour”,
American Economic Review, No. 53 (5), 1032-1057
Committee of European Banking Supervisors
(2008)
“Report on issues regarding the valuation of complex
and illiquid financial instruments”, June
130
Banque de France • Financial Stability Review • No. 12 – Valuation and financial stability • October 2008